LSE Research Online

[Pages:14]LSE Research Online

Book Section

Sandra Jovchelovitch and Martin W. Bauer

Narrative interviewing

Originally published as: Bauer, Martin W.; Gaskell, G.; Eds., Qualitative researching with text, image and sound : a practical handbook. London, England: SAGE Publications, 2000. ? 2000 SAGE Publications.

You may cite this version as: Jovchelovitch, Sandra; Bauer, Martin W. (2000). Narrative interviewing [online]. London: LSE Research Online. Available at: Available in LSE Research Online: August 2007

LSE has developed LSE Research Online so that users may access research output of the School. Copyright ? and Moral Rights for the papers on this site are retained by the individual authors and/or other copyright owners. Users may download and/or print one copy of any article(s) in LSE Research Online to facilitate their private study or for non-commercial research. You may not engage in further distribution of the material or use it for any profit-making activities or any commercial gain. You may freely distribute the URL () of the LSE Research Online website. This document is the author's final manuscript version of the book section, incorporating any revisions agreed during the review process. Some differences between this version and the publisher's version may remain. You are advised to consult the publisher's version if you wish to cite from it.

Contact LSE Research Online at: Library.Researchonline@lse.ac.uk

Narrative Interviewing

Sandra Jovchelovitch and Martin W. Bauer

The study of narratives has gained a new momentum in recent years. This renewed interest in an old topic - concern with narratives and narrativity goes back to Aristotle's Poetics - is related to increasing awareness of the role story-telling plays in shaping social phenomena. In the wake of this new awareness, narratives have become a widespread research method in the social sciences. Discussion about narratives, however, goes far beyond their use as a method of inquiry. Narrative as a discursive form, narratives as history, and narratives as life stories and societal stories have been approached by cultural and literary theorists, linguists, philosophers of history, psychologists and anthropologists.

This chapter addresses the use of narratives in social inquiry by discussing some elements of narrative theory and introducing the narrative interview as a specific technique of data collection, in particular in the format systematized by Sch?tze (1977; 1983; 1992). In what follows we introduce conceptual issues related to narratives, and present the narrative interview as a method of data elicitation, discussing in detail the procedure, indication for its use and potential problems associated with this technique. We conclude with a discussion of the thorny epistemological problem of what, in fact, narratives tell us.

Conceptual issues

There is no human experience that cannot be expressed in the form of a narrative. As Roland Barthes pointed out:

Narrative is present in myth, legend, fable, tale, novella, epic, history, tragedy, drama, comedy, mime, painting (think of Carpaccio's Saint Ursula), stained-glass windows, cinema, comics, news items, conversation. Moreover, under this almost infinite diversity of forms, narrative is present in every age, in every place, in every society; it begins with the very history of mankind and there nowhere is nor has been a people without narrative. . . Caring nothing for the division between good and bad literature, narrative is international, trans historical, transcultural: it is simply there, like life itself. (1993: 251-2)

Indeed, narratives are infinite in their variety, and we find them everywhere. There seems to be in all forms of human life a need to tell; storytelling is an elementary form of human communication and, independently of stratified language performance, it is a universal competence. By telling, people recall what has happened, put experience into sequence, find possible explanations for it, and play with the chain of events that shapes individual and social life. Story-telling involves intentional states that alleviate, or at least make familiar, events and feelings that confront ordinary everyday life.

Communities, social groups and subcultures tell stories with words and meanings that are specific to their experience and way of life. The lexicon of a social group constitutes its perspective on the world, and it is assumed that narrations preserve particular perspectives in a more genuine form. Story-telling is a skill relatively independent of education and language competence; while the latter is unequally distributed in any population, the competence to tell stories is not, or at least is less so. An event can be rendered either in general terms or in indexical terms. Indexical means that reference is made to concrete events in place and time. Narrations are rich in indexical statements, (a) because they refer to personal experience, and (b) because they tend to be detailed with a focus on events and actions; The structure of a narration is similar to the structure of orientation for action: a context is given; the events are sequential and end at a particular point; the narration includes a kind of evaluation of the outcome. Situation, goalsetting, planning, and evaluation of outcomes are constituents of human goal-directed actions. The narration reconstructs actions and context in the most adequate way: it reveals place, time, motivation and the actor's symbolic system of orientations (Sch?tze, 1977; Bruner, 1990).

The act of telling a story is fairly simple. According to Ricoeur (1980), someone puts a number of actions and experiences into a sequence. These are the actions of a number of characters, and these characters act out situations that change. The changes bring to light elements of the situation and of the characters that were previously implicit. In doing so, they call thinking, or for action, or for both. Storytelling comprises two dimensions: the chronological dimension, which refers to the narrative as a sequence of episodes, and the non-chronological, which involves the construction of a whole from successive events, or the configuration of a plot. The plot is crucial to the constitution of a narrative structure. It is through the plot that individual units (or smaller stories within the big story) in the narrative acquire meaning. Therefore a narrative is not just a listing of events, but an attempt to link them both in time and in meaning. If we consider events in isolation they appear to us as simple propositions that describe independent happenings. But if they are composed into a story, the ways in which they are related allow for the meaning-production operation of the plot. It is the plot that gives coherence and meaning to the narrative, as well as providing the context in which we understand each of the events, actors, descriptions, goals, morals and relationships that usually form a story.

Plots operate through specific functions that serve to compose and configure various events into a narrative. First, it is the plot of a narrative that defines the temporal range that marks the beginning and end of a story. We know that human life, and the vast majority of social phenomena, flow without precise beginnings and ends. Yet in order to make sense of life's, events and understand what is going on, it is important to demarcate beginnings and ends. Secondly, the plot provides criteria for the selection of events to be included in the narrative, for how these events are ordered into an unfolding sequence until the conclusion of the story, and for the clarification of the implicit meanings that events have as contributors to the narrative as a whole. Deciding what is to be told and what is not, and what should be told first, are operations related to the meaning that a plot gives to a narrative.

In this sense, narratives live beyond the sentences and events that form them; structurally, narratives share the characteristics of the sentence without ever being reducible to the simple sum of its sentences or forming events. In the same vein, meaning is not at the 'end' of the narrative; it permeates the whole story. Thus to understand the narrative is not merely to follow the chronological sequence of events that are unfolded by the story-teller: it is also to recognize its non-chronological dimension expressed by the functions and meanings of the plot.

The narrative interview

The narrative interview (henceforth, NI) envisages a setting that encourages and stimulates an interviewee (who in NI is called an 'informant') to tell a story about some significant event in their life and social context. The technique derives its label from the Latin word narrare, to report, to tell a story. In an unpublished manuscript, Sch?tze (1977) has suggested a systematization of this technique. Its basic idea is to reconstruct social events from the perspective of informants as directly as possible. To date, we have used narrative interviews to reconstruct informants' perspectives in two studies: first, to reconstruct actors' perspectives in a controversial software development project in a corporate context (Bauer, 1991; 1996; 1997); and secondly, to investigate representations of public life in Brazil (Jovchelovitch, 2000). Positive experiences encourage us to recommend the technique, and to make Sch?tze's systematization available in the English language with some elaboration.

This particular version of narrative interviewing has not been accessible in English, although writings about narratives abound in different versions. Most writings about 'narratives' have an analytic focus, stressing the structural characteristics and philosophical significance of narratives (Riesman, 1993; Barthes, 1993; Bruner, 1990; Mitchell; 1980; Johnson and Mandler, 1980; Kintsch and VanDijk, 1978; Propp, 1928). The strength of Sch?tze's suggestion is a systematic proposal for eliciting narratives for the purposes of social research. Sch?tze's manuscript of 1977 remains unpublished; it spread widely as grey literature and became the focus of a veritable method community in Germany

during the 1980s. The original idea developed out of a research project on power structures in local communities.

Narrative as a self-generating schema: (once upon a time)

Story-telling seems to follow universal rules that guide the process of story production. Sch?tze (1977) describes as the 'inherent demands of narration' (Zugzwaenge des Erzaehlens) what others have called 'story schema', 'narrative convention' or 'story grammar' (Johnson and Mandler, 1980; Kintsch and VanDijk, 1978; Labov, 1972). A schema structures a semi-autonomous process activated by a predetermined situation. A narration is thus elicited on the basis of particular clues, and, once the informant has started, storytelling will sustain a flow of narration drawing on underlying tacit rules.

Story-telling follows a self-generating schema with three main characteristic as follows.

DETAILED TEXTURE

This refers to the need to give detailed information in order to

account plausibly for the transition from one event to another. The narrator tends to give as much detail

of events as is necessary to make the transition between them plausible. This is done taking the listener

into account: the story has to be plausible for an audience, otherwise it is no story. The less the listener

knows, the more detail will be given. Storytelling is close to events. It will account for time, place,

motives, points of orientation, plans, strategies and abilities.

RELEVANCE FIXATION

The story-teller reports those features of the event that are relevant

according to his or her perspective on the world. The account of events is necessarily selective. It

unfolds around thematic centres that reflect what the narrator considers relevant. These themes

represent his or her relevance structure.

CLOSING OF THE GESTALT A core event mentioned in the narration has to be reported completely, with a beginning, a middle and en end. The end can be the present, if the actual events are not yet finished. This threefold structure of a closure makes the story flow, once it has started: the beginning tends towards the middle, and the middle tends towards the end.

Beyond the question-answer schema

The narrative interview is classified as a qualitative research method (Lamnek, 1989; Hatch and Wisniewski, 1995; Riesman, 1993; Flick, 1998). It is considered a form of unstructured, in-depth interview with specific features. Conceptually, the idea of narrative interviewing is motivated by a critique of the question-response schema of most interviews. In the question-response mode the interviewer is imposing structures in a three fold-sense: (a) by selecting the theme and the topics, (b) by ordering the questions and (c) by wording the questions in his or her language.

To elicit a less imposed and therefore more 'valid' rendering of the informant's perspective, the influence of the interviewer should be minimal, and the setting should be arranged to achieve this minimizing of the interviewer's influence. The rules of engagement of the NI restrict the interviewer. The NI goes further than any other interview method in avoiding pre-structuring the interview. It is the most notable attempt to go beyond the question-answer type of interview. It uses a specific type of everyday communication, namely story-telling and listening, to reach this objective.

The narration schema substitutes the question-answer schema that defines most interview situations. The underlying presupposition is that the perspective of the interviewee is best revealed in stories where the informant is using his or her own spontaneous language in the narration of events. However, it would be naive to claim that the narration is without structure. A narrative is formally structured; as we pointed out above, narration follows a self-generating schema. Whoever tells a good story complies with the basic rules of story-telling. Here the paradox of narration arises: it is the constraints of the tacit rules that liberate the story-telling.

The technique is sensitive to two basic elements of interviewing, as pointed out by Farr (1982): it contrasts different perspectives, and takes seriously the idea that language, as the medium of exchange, is not neutral but constitutes a particular worldview. Appreciating difference in perspectives, which can be either between interviewer and informant or between different informants, is central to the technique. The interviewer is advised to carefully avoid imposing any form of language not used by the informant during the interview.

The elicitation technique

Table 4.1 Basic phases of the narrative interview

Phases Preparation 1 Initiation 2 Main narration 3 Questioning phase

4 Concluding talk

Rules Exploring the field Formulating exmanent questions Formulating initial topic for narration Using visual aids No interruptions Only non-verbal encouragement to continue story-telling Wait for the coda Only 'What happend then?' No opinion and attitude questions No arguing on contradictions No why-questions Exmanent into immanent questions Stop recording Why-questions allowed Memory protocol immediately after interview

The NI as an interview technique consists of a number of rules on: how to activate the story schema; how to elicit narrations from the informants; and how, once the narrative has started, to keep the narration going by mobilizing its self-generating schema. The story develops out of factual events, anticipated interest of the audience, and formal manipulations within the setting. The rules that follow are a mixture of Sch?tze's proposal and our own elaboration.

Table 4.1 summarizes the basic concept of the NI and its rules of procedure. The narrative interview is conducted over four phases: it starts with the initiation, moves through the narration and the questioning phase, and ends with the concluding talk phase. For each one of the phases, a number of rules are suggested. The function of these rules is not so much to encourage blind adherence, but to offer guidance and orientation for the interviewer in order to elicit rich narration on a topic of interest, and to avoid the pitfalls of the question-answer schema of interviewing. Following these rules is likely to lead to a non-threatening situation, and maintains the informant's willingness to tell a story about significant events.

Preparing the interview

Preparing for the NI takes time. A preliminary understanding of the main event is necessary both to make clear the gaps that the NI is to fill, and to achieve a cogent formulation of the initial central topic designed to trigger a self-sustainable narration. Primarily, the researcher needs to make themselves familiar with the field under study. This may involve making preliminary enquiries, reading documents and taking note of rumours and informal accounts of any particular event. Based on these preliminary inquiries, and on their own interests, the researcher draws up a list of `exmanent' questions. Exmanent

issues reflect the interests of the researcher and their formulations and language. From this, we distinguish `immanent' issues: the themes, topics and accounts of events that appear during the narration by the informant. Exmanent and immanent issues may overlap totally, partially or not at all. The crucial point of the exercise is to translate exmanent questions into immanent ones by anchoring exmanent issues in the narration and by applying nothing but the informant's own language. In the course of the interview, the attention of the interviewer should be focused on the immanent issues, on making notes of the language used, and on preparing for further questions at the appropriate time.

Phase 1: initiation

The context of the investigation is explained in broad terms to the informant. He or she must be asked for permission to record the interview on tape. Recording is important to support a proper analysis later. Then the procedure of the NI is briefly explained to the informant: uninterrupted story-telling, questioning phase and so on. In the preparation phase for the NI, an initial topic of narration has been identified. It should be noted that the initial topic represents the interests of the interviewer. To support the introduction of the initial topic visual aids may be used. A time-line, schematically representing the beginning and the end of the event in question, is a possible example. The narrator in this case would face the task of segmenting the time between the beginning and the end of the story.

The introduction of the central topic of the NI should trigger the process of narration. Experience shows that in order to elicit a sustainable story, several rules may be used as guidelines to formulate the initial topic:

? The initial topic needs to be experiential to the informant. This will ensure his or her interest, and a narration rich in detail.

? The initial topic must be of personal and of social or communal significance. ? The informant's interest and investment in the topic should not be mentioned. This is to avoid

taking positions and role playing from the very beginning. ? The topic shall be broad to allow the informant to develop a long story which, from an initial

state of affairs, through past events, leads to the present situation. ? Avoid indexical formulations. Do not refer to dates, names or places. These should be

introduced only by the informant as part of his or her relevance structure.

Phase 2: main narration

When the narration starts, it must not be interrupted until there is a clear coda, meaning that the interviewee pauses and signals the end of the story. During the narration, the interviewer abstains from any comment other than non-verbal signals of attentive listening and explicit encouragement to continue the narration. The interviewer may, however, take occasional notes for later questioning, if this does not interfere with the narration.

Restrict yourself to active listening, non-verbal or paralinguistic support and showing interest ('Hmm, 'yes', 'I see'). While listening, develop, in your mind or on paper, the questions for the next phase of the interview.

When the informant marks the coda at the end of the story, probe for anything else: 'is this all you want to tell me?' or 'is there anything else you want to say?'

Phase 3: questioning phase

As the narration comes to a 'natural' end, the interviewer opens the questioning phase. This is the moment when the attentive listening by the interviewer bears its fruits. The exmanent questions of the interviewer are translated into immanent questions using the language of the informant to complete the gaps in the study. The questioning phase should not start unless the interviewer has sufficiently probed the end of the main narrative. In the questioning phase, three basic rules apply:

? Do not ask why-questions; ask only questions concerning events like 'what happened before/after/then?' Do not directly ask about opinions, attitudes or causes as this invites justifications and rationalizations. Every narrative will include the latter; however, it is important not to probe them, but to see them occurring spontaneously.

? Ask only immanent questions, using only the words of the informant. Questions refer both to events mentioned in the story and to topics of the research project. Translate exmanent questions into immanent questions.

? To avoid a climate of cross-examination, do not point to contradictions in the narrative. This is again a precaution against probing rationalization beyond that which occurs spontaneously.

The questioning phase is meant to elicit new and additional material beyond the self-generating schema of the story. The interviewer asks for further 'concrete texture' and 'closing of the Gestalt' by staying within the rules.

Phases 1, 2 and 3 are recorded for verbatim transcription with the consent of the informants.

Phase 4: concluding talk

At the end of the interview, as the tape recorder is switched off, interesting discussions often develop in the form of small-talk. Talking in a relaxed mood after the 'show' often throws light on the more formal accounts given during the narration. This contextual information proves in many cases to be very important for the interpretation of the data, and it can be crucial for a contextual interpretation of the informants' accounts.

During this phase. the interviewer may use why-questions. This may be an entry point for the analysis later, when the theories and explanations that the story-tellers hold about themselves ('eigentheories') become a focus of analysis. Furthermore, in this last phase the interviewer may also be in a position to rate the level of (mis)trust they command in the eye of the informant, which is important information for the interpretation of the narration in its context.

In order not to miss this important information, it is advisable to have a notebook or a prepared form for summarizing the contents of the small-talk in a memory protocol immediately after the interview. If one organizes a series of NIs, it is worth planning for time between the interviews for writing up the small-talk and other impressions.

Strengths and weaknesses of the narrative interview

Researchers using the narrative interview have pointed to two main problems of the technique: (a) the uncontrollable expectations of the informants, which raise doubts about the strong claim of nondirectivity of the NI, and (b) the often unrealistic role and rule requirements of its procedures.

Uncontrollable expectations in the interview

The interviewer aims to obtain a complete narration of events out of every interview, which expresses one specific perspective. They therefore pose as someone who knows nothing or very little about the story being told, and who has no particular interests related to it. Every participant, however, will make hypotheses about what the interviewer wants to hear and what they probably already know. Informants generally assume that the interviewer does know something about the story, and that they do not talk about what they know because they take it for granted. It is highly problematic to stage a `pretend play' of naivety, especially over a series of interviews where the informant knows that he or she is not the first to be interviewed.

As noted above, every informant will make hypotheses about what the interviewer wants to hear. The interviewer must thus be sensitive to the fact that the story they obtain is to some degree

strategic communication, that is, it is a purposeful account either to please the interviewer, or to make a particular point within a complex political context that may be at stake. It may be difficult, if not impossible, to obtain a narrative from a politician that is not a strategic communication. The informant might try to defend himself or herself in a conflict, or might put himself or herself in a positive light with regard to the events.

The interpretation of the NI must take into account such possible circumstances, which are unavoidable in the very situation of the interview. The narration in an NI is a function of the whole situation, and it has to be interpreted in the light of the situation of the study, the assumed strategy of the narrator and the expectations that the informant attributes to the interviewer. Independently of what the interviewer says, the informant may suspect a hidden agenda. Alternatively, the informant may trust the interviewer, not assume a hidden agenda, and render an authentic narrative of events, but may at the same time transform the interview into an arena to advance his or her perspective for purposes wider than the research agenda.

The texture of the narrative will depend to a large extent on the pre-knowledge that the informant attributes to the interviewer. To play ignorant may be an unrealistic role requirement on the interviewer. Each interview demands that the interviewer presents themselves as ignorant, while in fact, their actual knowledge is increasing from one interview to the next. The credibility of this pretence has limits, and the knowledge of the interviewer cannot be hidden for long.

Under these circumstances, Witzel (1982) is sceptical about the claim that the relevance structures of the informants are revealed by narration. Any conversation is guided by expectations of expectations. Even in cases where the interviewer abstains from framing questions and answers, the active informant will tell her story to please or to frustrate the interviewer, or to use the interviewer for purposes beyond the interviewer's control. In all cases, the informant's relevance structures may remain hidden. The narration reflects the interpretation of the interview situation. Strategic storytelling cannot be ruled out.

Unrealistic rules

The rules of the NI are formulated to guide the interviewer. They are set up to preserve the informant's willingness to narrate some controversial events and problems under study. The main question is whether these rules are as helpful as they are well intended. Again, Witzel (1982) doubts that the prescribed format of the 'initial topic' is, in fact, suitable for every informant. The interviewer presents themselves as if they knew nothing about the topic under study. Informants might perceive this attitude as a trick, and this perception will interfere with their cooperation.

The way the interviewer initiates the interview co-determines the quality of the narration. This puts too much focus on the beginning of the interview. The narration is likely to be an outcome of the way the interviewer comports him or herself. The initiation phase is difficult to standardize, and relies totally on the social skills of the interviewer. This sensitivity of the method to the initial moment may be a cause of anxiety and stress for the interviewer. This may pose a difficulty for applying the NI technique in a research project with several interviewers who have different levels of skill. Another point of criticism refers to the fact that the rules of the NI technique were developed in a specific field study, which dealt with local politics and biographical research. The rules express suggestions for coping with the problem of interaction in these specific studies, and might not work as intended in other circumstances. This is an empirical problem that must be studied by applying the NI in different circumstances. However, little methodological research beyond a description or a general critique of the technique has been conducted.

The rules of the narrative interview define an ideal-typical procedure which may only rarely be accomplished. They serve as a standard of aspiration. In practice, the NI often requires a compromise between narrative and questioning. The narratives reveal the diverse perspectives of the informants on events and on themselves, while standardized questions enable us to make direct comparisons across various interviews on the same issue. Furthermore, an interview may go through several sequences of narration and subsequent questioning. The iteration of narration and questioning may occasionally blur

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download